Training and match load quantification in soccer is a routine practice (1–4). Load quantification can be expressed by the locomotor/mechanical demands or physical demands that can associated with external load. Usually, global navigation satellite system (GNSS) metrics, such as distances covered at various running speeds or accelerations, are used to quantify external load (2, 5).

External training load may vary for players according to the number of training/match sessions per week/microcycle due to specific aims of training sessions and the periodisation/planning strategies of the coach (6), or even changing the coach through the season (7). For instance, a recent systematic review in professional male soccer revealed that three to six training sessions were performed per week (8), highlighting that different strategies were applied. Other factors that impact load could be related to the use of specific drills and games during training (e.g., small-sided games, long sprints, repeated sprints, interval training, and medium- to large-sized games) (9). Regarding match-play, load quantification can vary due to the dynamics of matches and contextual factors (10–13).

Previous studies have suggested that the number of training sessions affects the training load in male soccer (14–16). However, it is relevant to mention that different methods for load quantification were used. Anderson et al. analysed the load of each training and match session of three microcycles: a) one with four training sessions and one match; b) another with four training sessions and two matches; and c) with two training sessions and three matches (14). Oliveira et al. analysed the load of each training and match session and the weekly average values of five different microcycles in which all had four training sessions, although the number of matches varied from one to three (15). Nobari et al. analysed the weekly mean values between all microcycles and the accumulated mean values of all weeks included for each microcycle in which the number of training sessions varied from two to six and the number of matches varied from one to two (16).

A method to clarify the understanding of the load induced by training and competitive match-play is to calculate the training/match ratio (TMr) (6, 17, 18). This ratio is calculated by dividing the accumulated weekly load by the match load (6). If the TMr provides a value lower than one, it suggests that the accumulated load of the week is lower than the match load; a value above one suggests that the training load is greater than the match load (6).

Despite the potential practical implications of such analyses, there are a limited number of investigations in male soccer players that examined this ratio with a varying number of training sessions (6, 18) and to the best of authors knowledge, there is no research in female soccer players, although one study analysed the density of training in relation to match-play (19). Clemente et al. compared the TMr of different external load measures between weeks with three, four, and five training sessions/week. The study showed that weeks with five training sessions had higher values for all external load ratios than weeks with three or four training sessions. Additionally, high-speed running (HSR) and sprint distance (SPD) measures presented substantially lower ratios than other variables such as total running distance, accelerations (ACC), decelerations (DEC), and player load (6). Oliveira et al. also compared the TMr of different external load measures between weeks with three, four, five and six training sessions/week. The study showed that weeks with five and six training session presented higher values than weeks with three and four training sessions. The study also demonstrated that weeks with three and four training session had some metrics with TMr values lower than one (e.g, HSR, ACC and DEC) (18). Furthermore, Olaizola et al. compared external load measures of training sessions utilising official match-play as a reference in women's soccer. Key findings revealed that none of the training sessions obtained higher values than matches for all examined measures (19).

Therefore, this study aimed to: (i) compare accumulated and relative accumulated load between three and four training sessions plus a match weeks across two consecutive seasons (2019–20 and 2020–2021) in elite female soccer players, and (ii) to compare the training/match ratio (TMr) of external load. Based on previous studies (6, 18), it was hypothesised that the season with more training sessions will display higher loads.

Table 1 presents the comparisons for accumulated load demands (match data included). Except for ACC4 and DEC4, all variables showed meaningful differences (p < 0.05) with moderate to very large effect sizes.

Table 2 presents comparisons for all TMr. Apart for ACC4r and DEC4r, all variables showed meaningful differences (p < 0.001) with moderate to very large effect sizes and only ACC3r showed a small effect.

Figure 1 shows the accumulative weekly training load, match load and training/match ratios per microcycle.

Figure 2 shows match-day minus description of the different metrics in percentages of the 3dW, while Figure 3 presents 4dW.

Table 3 presents comparisons for relative accumulated load demands. Only total distance and HSR showed meaningful differences (p < 0.05) with large to very large effect sizes.

The aims of this study were to: (i) compare accumulated and relative accumulated load between three and four training session plus a match across two consecutive seasons (2019–20 and 2020–2021) in elite female soccer players, and (ii) compare the training/match ratio (TMr) of external load.

Training load monitoring has become a relevant aspect between research and practice to control training and match demands in team sports such as soccer. Comparing accumulated training and match load in weeks with varying numbers of training sessions can help support the understanding of whether an increase in the different load metrics is evident (32). The use of GNSS technology provides valuable information on external load in terms of the volume and intensity of each training session and match (33). Previous research has shown that over or under exposure to some external loading metrics, such as HSR and ACC/DEC, may increase the risk of non-contact injuries (34, 35). Similarly, higher accumulated total distance has been associated with injury in female soccer players (36).

The present study findings revealed that when comparing accumulated load demands between three and four training sessions across two consecutive seasons (match data included) all variables, except for ACC >4 m.s⁻² and DEC <4 m.s⁻², showed meaningful differences (p < 0.05) with moderate to very large effect sizes.

Previous research suggested that match-play represents the greatest physiological stimulus and represents the primary performance outcome (37, 38). Nonetheless, nearly 80% of the weekly training load resulted from the training sessions, whereas about 20% came from the match-play (39, 40). Understanding the cumulative effect of training is essential to guide the individual athlete's performance (5, 41). To optimise athletic performance, training sessions should be prescribed to suit an individual athlete's physical characteristics. However, in team sports this is complex by nature and complicated to apply (42). In the Portuguese female soccer league, training sessions are often conducted three times per week, which reduces the likelihood that players are performing training appropriate to the demands of the game (43). Quantification of training/match load represents an important procedure to adjust the training stimuli provided to players to adequately prepare for match demands (5, 6, 18).

The differences in external load between microcycles with three and four training sessions plus a match in elite female soccer players across two competitive seasons can potentially be partly explained by several factors, including training periodization, recovery strategies, match intensity, and player fitness levels. When analysing the average duration of the training sessions plus the match, in the two competitive microcycles of three and four sessions, the values ​​were similar, 83.15 and 88.34 min, respectively. Regardless of the number of weekly sessions, the accumulated duration appears to maintain the same temporal pattern. For total distance, there was a significant decrease in the distance covered in the four-sessions microcycle. This may possibly be due to the coach opting for a weekly training structure with more sessions dedicated to tactical and skills training, although the analysis of the training content was beyond the scope of this research, in the season with four weekly training sessions. However, the total distance covered in the microcycles with three sessions is similar to those in one of the few studies that analysed training load in women's soccer (44). As reported previously, during a competitive microcycle, the workload may be influenced by contextual factors such as the length of the microcycle (45, 46). Furthermore, Gualtieri et al. established that practitioners adjusted the microcycle schedule based on specific microcycle lengths (i.e., 3-, 4-, 5-days), reducing muscular impact (i.e., ACC and DEC) and using the 5-day microcycle as an opportunity for recovery (47). However, in the present study, the absolute accumulated load of ACC and DEC was higher in the 4dW. When comparing the number of ACC between microcycles, the 4dW reported higher values, except for ACC4. The same pattern was observed for DEC. Posse-Álvarez et al. reported that longer microcycles (four or five days) showed significantly greater ACC and DEC than shorter microcycles (two days) (46). Perhaps these differences in mechanical load could be related to shorter training session durations in the four-session microcycle (∼15 min less than the three-session week).

It is important to note that in the present study, the match number was higher in the three-session microcycle (15 vs. 10 matches), which may have contributed to the differences in total distance between the microcycles. Despite using the average values for analysis, there was a higher standard deviation for the total distance metric in the second season than in the first. Other justifications may be associated with the highest intensity displayed in the 4dW while all other metrics were higher when compared with 3dW. In other words, the higher intensity and number of matches performed in the second season (4dW) contributed to a lower total distance covered when compared with the first season (3dW). Additionally, the increase in the number of training days and the loading pattern of the 4dW compared with the 3dW may further affect physical performance during match-play during the second season. For example, the present study did not examine the differences in training sessions (e.g., MD-5, MD-4, MD-3, MD-2) which may explain the lower total distances in the second season, while sprint distance (e.g., > 18 km.h⁻¹) that is known to impact match-play (48, 49), was not reported, since the present study used a different threshold (>15 km.h⁻¹) of HSR (27).

The distances covered at HSR in the two microcycles were significantly different, with a greater distance covered in the 4dW (1820.35 vs. 4,044.77 m). However, teams often manage loads by adjusting intensity in 4dW to avoid excessive fatigue before matches (50). The 3dW often allow enhanced recovery, potentially benefiting match performance, especially in congested periods. In the study by Mara et al., a typical weekly in-season training structure consisted of one match followed by two recovery days, one conditioning session, one skill session, and two tactical sessions, with a distance of 1,027 m reported during the early season (44). However, other authors reported that female players covered 718 m, which was around ∼7% of the match total distance (51). However, attention should be given to HSR and SPD distances as important indicators of match physical performance (37, 51).

Overall, this study demonstrated that the external load of training microcycles varies depending on the length of the microcycles in professional female soccer, and these results are also supported by a recent study (46). Considering that the quantification of training/match load represents an important procedure for adjusting training stimuli provided to players to adequately cope with the demands of the match (33, 38), in the present study, the training/match ratios were analysed. Thus, from the analysis of the TMr, although the relationship between accumulated weekly training load and match demands varies according to the nature of the external load measure, it was observed that all metrics showed meaningful differences, except for ACC4 and DEC4 (without significant results). Oliva-Lozano et al. reported that the length of the microcycle had a significant effect on the load, not only in the volume but also in the intensity, except for relatively high ACC and DEC (45). This study found that the TMr were higher in the 4dW, while previous studies showed that the length of the microcycle may have an impact on the workload of soccer players (6, 14–16, 18, 33, 52), which supports the present findings. As an interpretation guidance, when the TMr is higher than one, it suggests that the weekly training load is greater than the match load. When the value is lower than one, it reports that the weekly training load is lower than the match load. When the value is equal to one, it means that both are equivalent (6). In the present study, only HSRr in the 3dW presented a close value to one (1.11 A.U.), suggesting that HSR was slightly higher than during a match.

Finally, the relative accumulated load calculated for each measure (absolute accumulated load/absolute accumulated duration of the microcycle) showed significant differences, although only in total distance (63.80 vs. 35.17 m/min) and HSR distance (5.46 vs. 9.27 m/min). The results from recent studies have shown that the training load increased with longer microcycles (6, 18, 53). This is a questionable issue, as our training load data did not consistently show this tendency, which could be attributed to variations in the training methodologies used by the coaches (45).

This study addressed a general overview of load monitoring in professional female soccer players. Considering that the training load can be influenced by the type of microcyle, player's starting status, playing positions, training mode, and other contextual factors, the comparison of the two seasons indicates that accumulated load and TMr increase with the number of weekly training sessions. Understanding seasonal competitive training/match load variations and the relationships between measures seems important to define the most appropriate monitoring strategy of external load during varying microcycle structures.

This study is not without limitations that could inform future research. For instance, it analysed only a single professional female soccer team from a Portuguese league, thus highlighting the potential value of conducting a multi-club study across different leagues and countries. Still, caution should be taken when interpreting the results. Additionally, future studies could incorporate internal load variables (e.g., RPE, mean heart rate) and contextual variables such as match result, opponents, and location since they have been shown to play some role in load distribution across weeks in female players (54).

This study showed that adding one training session per week increased accumulated load and TMr for several key variables, which suggest that in the first season, additional load may have been applied even in fewer training sessions. Specifically, when analysing accumulated load demands, more total distance was covered in 3dW than 4dW, while all others were higher in 4dW. Moreover, when considering TMr, all metrics were higher in 4dW. Still, no TMr lower than 1.0 were found. Lastly, when considering relative accumulated load demands, more total distance was covered in 3dW than 4dW while more HSR was covered in 4dW than 3dW. However, no other meaningful differences were found which reveals the importance on how analysing data can change the results interpretation.